My RAW Processing Workflow

Here's a rundown of the primary adjustments (tool-agnostic) that I make to most photos - in order and why I make them. It's a bit of a re-post, so I'm going to add lots of other detail and insights on each step. Please keep in mind that this is how I do it, but it may not be the best way to process your images. With that said, please feel free to ask me questions or critique my methods:

Notes on Post-Processing (PP)

(A) Do as much as you can in RAW as it's "non-destructive", Any adjustments made after this step will slowly (16-bit PSD or TIFF) or quickly (8-bit / JPEG) begin to rob the image of resolution and create banded skies, etc. Once your histogram looks like a comb instead of a rolling hill, you've overdone it.

(B) Working with RAW makes me think of my days as a (analog film) movie projectionist. Each time I started the projector, I would stand there and focus the lens. I would turn the knob until the image was clearly out-of-focus, and then turn it the other way until it was out-of-focus again, then make smaller turns until I got perfect focus. This is the way I try to adjust my photos. Too much of any one adjustment is bad, but not enough makes the photo appear lifeless and flat as well. The idea is to find the balance between the two, with an eye to underdoing it vs. overdoing it, which is all too easy to do.

(C) A calibrated monitor (D6500 for most purposes or D5000 for commercial printing) is nearly essential to editing. It's well worth the price of even the cheapest calibration tools from XRite, Colovision, and others to get accurate color and brightness. Also, do the processing in a dim room with no reflections on the screen for best results.

(D) The steps below assume the use of a lens correction module from DxO or PhotoShop/LR Adobe Camera RAW. The only exception to the use of the automated tools is usually the use vignetting. I often find myself turning the vignetting correction with some types of photos (portraits or close ups) as the natural vignetting can add a nice look and focus to the photo. I'm not a fan of adding vignettes, but the natural ones can look nice.

Workflow Steps for RAW Processing

1. Cropping (as needed) - I usually begin by cropping so the photo I'm working on is exactly the framing I want

Additional Thoughts: Like most people, I try to get this right in camera, but sometimes there's an errant leaf or something to clip. I also find myself using the rule of thirds, golden ratio, or other guides like having the subject look out of the frame and balancing the elements in the frame.

2. White balance - fix this first as it affects exposure and colors

Additional Thoughts: For interiors or mixed lighting, I often take a white balance card - usually the ColorChecker Passport, but even a sheet of paper will do so I have a reference in post. For nature shots, getting WB correct is often quite difficult because there are no neutral settings. I will typically start with Daylight, Shade, or Cloudy and play with the sliders until I achieve the WB I like. For pre- and post-sunrise shots, it can be extremely high, going into the 9000s, but once the sun comes up, Daylight generally gives the best/closest results. I like to leave a bit of the golden color vs. going neutral, but that's just a matter of personal preference.

3. Global Contrast - again, this affects other adjustments - I typically add a touch to most shots, but more if there's flare, fog, or other things that have reduced contrast, unless that's the look I want to achieve.

Additional Thoughts: Back to the focus knob analogy, it's best to play with this one until you get a good balance. I always watch the shadows when I adjust this as it's easy to "crush" the blacks a bit if the contrast is too high.

4. Exposure - I expose the right for everything but studio lit stuff, so I usually bump down the midtones a bit with the exposure setting, or if there are no true blacks or whites, I adjust for the midtone level I want to achieve.

Additional Thoughts: It occurs to me that this is probably my least used adjustment because the control is too coarse for my tastes, i.e. it affects way too much of the image if used for anything other than minor adjustments.

5. Black & white points - usually with curves tool. I make sure to adjust with the over- or under-exposure shown so I don't go too far. If the shot is high or low key, I'm careful about doing too much or too little at either end. If the shot has no blacks or whites, I generally skip this step.

Additional Thoughts: For photos with good contrast, I'll often skip steps 3 & 4 and go straight to these adjustments as they will generally yield the best results for exposure and contrast.

6. Color - using HSL- if there is a color cast left after WB adjustments, I correct it here.

Additional Thoughts: Sometimes you have a green or blue color cast from fluorescent lighting that you can't seem to get out with WB alone. Knocking the saturation down for that shade will help with the color cast. The same thing can be used with mixed lighting as well, though that may require layer masks in PhotoShop, which is down the road from this step.

7. Vibrancy - with most lenses, I leave this alone or just add a touch, but some of my older lenses seemed to need a slight bump. If the light was flat or the High ISO robbed the color, I'll add it here.

Additional Thoughts: To clarify - with most of the Canon EF primes, color saturation is excellent, but some of the older zooms like the 24-70 f/2.8, I found the colors needed a bit of a boost. Shooting at high ISO (above 1600 or 3200 on most bodies) also robs the image of some of its richness the higher you go, so vibrancy helps add that back. Watch out with skies and greys as it adds a lot of blue and will quickly turn your sky an odd shade of blue/purple or turn a great blue heron's gray feathers rather blue.

8. Saturation - using HSL - If there's a specific color I want to emphasize (yellow flowers in a landscape, perhaps) or reduce (say reds in skintones), I'll use the HSL slider to add or reduce saturation for that specific color

Additional Thoughts: This is sort of like the exposure slider - it's a rather coarse adjusment. It works well

9. Local contrast - I'll typically add some local contrast to most shots, but will reduce it for portraits - this affects sharpness

Additional Thoughts: For fine detail like blades of grass, bricks in a building, hair, feathers, etc. this is a very useful adjustment, but watch out for moire and overly harsh transitions. It's a good idea to adjust this at 50% magnification which is close to what it will look like if printed.

10. Sharpness - this is usually my final global adjustment and I will tweak it depending on how sharp the focus/lens is, how high of ISO I used, and the subject matter.

Additional Thoughts: As I said in step 9, there's an old desktop publishing trick of adjusting sharpnes at 50% magnification. This is very close to how the sharpness will appear when it's printed. I'm not sure how Retina et al displays affect this, but it really does work.

11. Local edits - dust spots from the sensor, red eye, blemishes, etc. I try to do as much work with global changes during raw conversion and I prefer to do as much in camera as possible, but Photoshop is great for layer blends, retouching, masking, and much more if that's what I need.

Additional Thoughts: this will be the subject of future posts...but PhotoShop is my prime tool, with Topaz Labs Remask and Nik being the primary plug-ins that I use.
 
Besisika said:
Let me be the first to say it; "thank you". I appreciate the effort.
I will post something as well, when I have a time. Promise.

Meanwhile, if you can tell what portion you do in ACR/Lightroom and what you do in PS.
Thanks,
You're welcome and I hope that the information is helpful. As for the portion, I find myself doing 95-100% of my work in RAW for most photos with final 5% being simple layer blends for a digital ND grad, pano stitching, focus-stacking for macro, or very minor retouching/cloning or dodges/burns. The main exception is when I'm retouching portraits (80/20%) or building a composite of many photos (10/90%).
 
Upvote 0
climber said:
I would like to ask what actually is the diference if editing image in RAW converter (LR or Camera Raw) or 16 bit TIFF (or PSD) file in Photoshop. Workflow could be non-destructive in both cases, so this is not the reason.
I guess I should have qualified that comment :) With adjustment layers / masks in PS, all of the editing is non-destructive no matter what the format is, but if you simply adjust brightness, contrast, color, etc. without using a layer mask, it is a destructive edit. It is much less destructive in uncompressed 16-bit formats, but do it enough and the image will eventually suffer.

Ultimately, doing as much as you can in RAW is still best because you are manipulating the RAW data and the RAW converters like ACR and DxO actually use this data to interpolate the pixels differently for highlight recovery and such. Once the conversion is done, you are adjusting individual pixels which isn't going to yield as good of results for things like highlight and shadow recovery, noise reduction and other functions.

More than anything, though, doing it in RAW is much faster than creating adjustment layers.
 
Upvote 0
notapro said:
Thank you for your thoughtful post, Mackguyver. It's helpful and illuminating to be able to have a walk-through of one's workflow. Your idea on how working with RAW is analagous to adjusting a focus knob is also a fine way, I believe, to describe the experiential element of the process.
In some ways, it's a good way to look at a lot things in life, as well.
 
Upvote 0
mackguyver said:
climber said:
I would like to ask what actually is the diference if editing image in RAW converter (LR or Camera Raw) or 16 bit TIFF (or PSD) file in Photoshop. Workflow could be non-destructive in both cases, so this is not the reason.
I guess I should have qualified that comment :) With adjustment layers / masks in PS, all of the editing is non-destructive no matter what the format is, but if you simply adjust brightness, contrast, color, etc. without using a layer mask, it is a destructive edit. It is much less destructive in uncompressed 16-bit formats, but do it enough and the image will eventually suffer.

Ultimately, doing as much as you can in RAW is still best because you are manipulating the RAW data and the RAW converters like ACR and DxO actually use this data to interpolate the pixels differently for highlight recovery and such. Once the conversion is done, you are adjusting individual pixels which isn't going to yield as good of results for things like highlight and shadow recovery, noise reduction and other functions.

More than anything, though, doing it in RAW is much faster than creating adjustment layers.

I've heard that applying curves or levels in PS is somehow "bending" pixels and the image quality is suffering from this. Do you know, if this is the same if adjusting image only with blending modes. Let say, if duplicate layer and turn it to multiply blending mode which will make image darker.
 
Upvote 0
climber said:
I've heard that applying curves or levels in PS is somehow "bending" pixels and the image quality is suffering from this. Do you know, if this is the same if adjusting image only with blending modes. Let say, if duplicate layer and turn it to multiply blending mode which will make image darker.
If you're editing the image in 16-bit mode in the AdobeRGB or ProPhoto RGB colorspace, I don't think that the image quality will suffer. The key is using adjustment layers, which are non-destructive, vs. applying the curve directly to the layer. Duplicating the layer and using blending modes is also non-destructive, but I personally find it easier to work with the curves in the adjustment layer as it's much easier to fine-tune the adjustment. Maybe some posts on curves and non-destructive editing in PS would be a good idea. I'll see what I can put together over the weekend.
 
Upvote 0
mackguyver said:

Thank you for your detailed guide revealing your workflow.
(sorry for re-animating such a quite old thread).

I have a few questions regarding your thoughts and hints:

Contrast
ACR provides a global contrast slider as well as the possibility to modfiy the gradiation curve.
1) Isn't setting the curve to a S-shape basically the same as increasing the constrast by the contrast slider? Do you touch the gradiation curve during processing in LR/ACR?
2) this is more like a general ACR question. In my version (that came along with PS CS5), the "curves" tab has two tabs: "Parametric" and "point" (directly translated from German, so maybe the actual wording is different). While the first allows to adjust shadows, dark tones, light tones and highlights by the use if sliders, the second tab allows to add arbitrary points in the curve just as the gradiation tool in Photoshop. I assumed that it would make sense if the usage of either "parametric" or "point" would be mutually exclusive, but this is not the case. So, when changing values in both tabs, you actually apply two curves. When would you make use of this feature?

Color
As I am affected by a red-green blindness, I noticed that I have difficulties in detecting "false" colors in photos. (e.g., when some photographers here point out reddish skin tones in portraits or when they are discussing color accuracy of different cameras - in these cases I just scratch my head without seeing what they are talking about). Is, gerenally spoken, the main goal (aside some aesthetic intent) that the photo will reveal the actual, real colors of the scene? Or is the main goal to achieve colors, that just fit the scene in the best way (but might differ from what the scene actually looked like)?


And finally: This thread would be perfect with screenshots of your editing process, from the inital un-cooked image via intermediate steps to the final result.
Again, thank you for your effort and your willingness to share your knowledge.
 
Upvote 0
Modifying the global contrast slider in ACR is effectively manipulating the S curve, but does so using more sophisticated algorithms in an attempt to provide more balanced contrast. For photos needing it, I will frequently add a touch of global contrast but use the curves to fine tune areas like the low shadows and mid highlights. For example, in the shot below, the dynamic range was quite high with the sun risen but most of the trees still in shadow. Using the contrast slider washed out the sky so I used the curves and brought down the 220-240 brightness range to bring back the cloud detail. I also used the 10-20 range to pull the shadows up just a bit in the darkest areas to give the trees more definition.

_H2B1142_DxO-L.jpg


The difference between point (what I used above) and parametric is a longer answer, but I'll try to get something together.

As far as color, even those of us without color blindness have trouble with skintones, so don't feel bad! As far as the goal, I hate to say it, it depends. For my personal work, what you said is exactly right. I make every attempt to work on the image the day I shot it, at least to give it the initial look, which is how I remember seeing the scene. For my commercial work, I typically use a ColorChecker Passport to profile the scene, or the original painting or artwork, and use controlled lighting to replicate the colors as accurately as possible. For portraits, there used to be some cool PS plug ins that corrected skin tones with a pointer, but I think the company went out of or stopped making the plug in version. Most Canons shoot a bit warm, especially with the elderly, so toning down the reds a bit is generally a good idea. In the end, it really just depends on what you're trying to do, and your subject, and unless you're being paid to "get the colors right" or your skin tones don't look human, it really doesn't matter. If you and your client (if applicable) are happy, that's all the matters.

Finally - when I have the time, I will do some instructions side-by-side with screenshots in both DxO 10 (which is a great upgrade) and ACR.
 
Upvote 0
mackguyver said:
Modifying the global contrast slider in ACR is effectively manipulating the S curve, but does so using more sophisticated algorithms in an attempt to provide more balanced contrast. For photos needing it, I will frequently add a touch of global contrast but use the curves to fine tune areas like the low shadows and mid highlights.

Personally, I always found +contrast in ACR gives too harsh results for screen/web viewing, for a substitute I use a modified tone curve in combination with the blacks/whites slider. The only time I change +contrast is with the local brush (though often clarity works better) or -contrast globally for high dr shots.

Global +contrast (or gamma) might be required for printing or some output devices though, but I usually change this with ImageMagick on the ACR-processed shot.
 
Upvote 0
My workflow in Lr looks like this:

1. Lens corrections and aberrations
2. Cropping, tilting etc. if needed
3. WB
4. Brightness/Clarity (don't know the exact english word)
5. Gradation curve + white and black
6. Color Management
7. Sharpness
8. NR; every picture of mine has some NR applied, which I find to be absolutely crucial with the 7D

Interesting to learn how other photographers treat their images! :)
 
Upvote 0
mackguyver said:
Here's a rundown of the primary adjustments (tool-agnostic) that I make to most photos - in order and why I make them. ...

Thank you for including the why. Showing people the what is not enough to teach them. Your sharing your thought process really helps in understanding the what.

Teaching me what you do will allow me to copy you
Teaching why you did it will allow me to understand it and be able to apply it to my own style.
 
Upvote 0

DominoDude

Certified photon catcher
Feb 7, 2013
910
2
::1
AcutancePhotography said:
mackguyver said:
Here's a rundown of the primary adjustments (tool-agnostic) that I make to most photos - in order and why I make them. ...

Thank you for including the why. Showing people the what is not enough to teach them. Your sharing your thought process really helps in understanding the what.

Teaching me what you do will allow me to copy you
Teaching why you did it will allow me to understand it and be able to apply it to my own style.

*nods in agreement with AcutancePhotography*
A good teacher don't assume the students are idiots, even if they know very little about a subject. A good teacher don't try to show off how incredibly good they are themselves making everyone else feel like they are idiots. A good teacher nudges his/hers students along a rewarding path to excellence, and afterwards they should feel like they did it all by themselves and didn't even need the teacher.

This is star material by MackWhiteEagle! ;)
 
Upvote 0